airflow-mcd

What it is and what it does

airflow-mcd is an Apache Airflow provider package that bridges Airflow workflows with Monte Carlo's data observability platform. It supplies callbacks that automatically notify Monte Carlo when DAGs and tasks succeed, fail, retry, or miss SLAs; hooks to create authenticated sessions with the Monte Carlo API; and operators including a circuit breaker for data quality gates and a suite of dbt operators that integrate dbt artifact reporting. The package is built on pycarlo, Monte Carlo's Python SDK, and integrates with Airflow's standard connection and configuration patterns.

The provider is designed for teams running data pipelines in Airflow who want centralized data quality monitoring and incident detection without manually instrumenting each task. It supports both broad, all-in-one callback patterns and granular, explicit callback selection, allowing integration into existing DAGs with minimal refactoring. The circuit breaker operator can halt downstream tasks if data quality rules fail, enforcing data contracts before dependent work executes.

Use it for:

• Send Airflow task and DAG events (success, failure, retry, SLA miss) to Monte Carlo for centralized incident detection and alerting.
• Implement data quality gates in pipelines using SimpleCircuitBreakerOperator to block downstream tasks until custom SQL monitors pass.
• Automatically report dbt run artifacts (models, tests, lineage) to Monte Carlo via DbtRunOperator and related dbt command operators.
• Create custom operators extending BaseMcdOperator to implement domain-specific data observability logic using the pycarlo SDK.
• Monitor data pipeline health and SLA compliance across Airflow DAGs without adding custom webhook or logging code to each task.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Integrates Monte Carlo data observability into Apache Airflow workflows via callbacks, hooks, and operators that send DAG and task events to Monte Carlo and execute data quality checks.

Yes, if you run Airflow and use Monte Carlo for data observability. The package has low install friction, active maintenance, no known vulnerabilities, and permissive licensing. It is well-suited for teams wanting to integrate data quality monitoring into existing Airflow workflows. Requires Airflow 1.10.14+ and a Monte Carlo account with API credentials configured in Airflow connections.

Install

airflow-mcd on PyPI

pip install airflow-mcd

uv add airflow-mcd

poetry add airflow-mcd

Installing airflow-mcd

Before you install

Low install friction with a pure-Python wheel distribution. Marked as active maintenance with a recent release. Requires Apache Airflow 1.10.14 or greater as a peer dependency, which is a standard Airflow ecosystem constraint.

License in practice

Licensed under Apache 2.0 (permissive), allowing commercial use, modification, and distribution with minimal restrictions—standard for Airflow providers.

Quickstart

pip install airflow-mcd

from airflow_mcd.callbacks import mcd_callbacks
from airflow import DAG

dag = DAG('my_dag', **mcd_callbacks.dag_callbacks)

Requires Apache Airflow 1.10.14 or greater to be installed separately; Python 3.7 or greater.

Verify before relying

• Whether pycarlo SDK version constraints or compatibility issues exist beyond what the fact sheet states.
• Specific Monte Carlo API authentication setup requirements and whether connection configuration is well-documented.
• Performance impact of callbacks and circuit breaker operators on large DAGs or frequent task execution.

Package facts

Evidence: airflow_mcd-0.3.12-py3-none-any.whl

Tags

More Build Tools packages